Academia.eduAcademia.edu

Figure 4 – uploaded by Floriano De Rango

See full PDFdownloadDownload figure
results with a “default” configuration of AODV model example of Omnet++ in which mobile nodes are equipped with omnidirectional antennas. As fol- lows it is shown the result related to the received packets when mobile nodes use IsotropicAntenna model: Figure 8: Received Packets Omnet++ Omni Configura- tion.

Figure 8 results with a “default” configuration of AODV model example of Omnet++ in which mobile nodes are equipped with omnidirectional antennas. As fol- lows it is shown the result related to the received packets when mobile nodes use IsotropicAntenna model: Figure 8: Received Packets Omnet++ Omni Configura- tion.

Related Figures (10)

Next, we created an algorithm for dynamic as- signment power for each host based on the main beam direction and so on transmission/reception direction position coordinates of nodes. For a given node that produces a radiation pattern forming a beam with main lobe angle @o it will transmit/receive with a power value that will be the maximum or a fraction of this according to radiation pattern form. The proposed algorithm takes into account also of the sidelobes amplitude of the pattern, so it has implemented a classification level of side lobes based on transmission/reception direction angle val- ue. In Fig. 4 it is shown a logical block schema that illustrates the main steps and modules about modifi- cations accomplished on implementation level basis: Rectangle forms contain modules or class names; rounded rectangles contain class functions names or variables names; elliptics contain only output varia- bles names; double rounded elliptics contain only output variables of a key function, and it can be con- sidered as a final state of an automata. In Fig.4 we note that ScalarAnalogModelBase extends and so implements all functions of AnalogModelBase Class. In ScalarAnalogModelBase there is a function computeReceptionPower that assigns the power val-
Next, we created an algorithm for dynamic as- signment power for each host based on the main beam direction and so on transmission/reception direction position coordinates of nodes. For a given node that produces a radiation pattern forming a beam with main lobe angle @o it will transmit/receive with a power value that will be the maximum or a fraction of this according to radiation pattern form. The proposed algorithm takes into account also of the sidelobes amplitude of the pattern, so it has implemented a classification level of side lobes based on transmission/reception direction angle val- ue. In Fig. 4 it is shown a logical block schema that illustrates the main steps and modules about modifi- cations accomplished on implementation level basis: Rectangle forms contain modules or class names; rounded rectangles contain class functions names or variables names; elliptics contain only output varia- bles names; double rounded elliptics contain only output variables of a key function, and it can be con- sidered as a final state of an automata. In Fig.4 we note that ScalarAnalogModelBase extends and so implements all functions of AnalogModelBase Class. In ScalarAnalogModelBase there is a function computeReceptionPower that assigns the power val-
A New Switched Beam Smart Antenna Model for Extending Inet Omnet++ Framework
A New Switched Beam Smart Antenna Model for Extending Inet Omnet++ Framework
Table 1: Setting of parameters used in simulations.
Table 1: Setting of parameters used in simulations.
Figure 7: Transmission directions matrix. As we can observe, distance between each radiat- ing element is a good set for avoiding Grating Lobes in radiation pattern; position of each node is ran- domly generated by Omnet++; for simplicity, only host 7 generates UDP traffic and sends it to host 0,1,3,5 that are receiver nodes. Because results de- pend on angular position and so on transmission direction of each node with respect to each other, firstly we created an utility module Mempos, to take into account the current position of the node and for generating a matrix that maintains the current posi- ion angle between all nodes, that is useful for eval- uating simulations results. As follows, we show the so called “transmission directions matrix” generated from Mempos module. It will be very helpful for results analysis:
Figure 7: Transmission directions matrix. As we can observe, distance between each radiat- ing element is a good set for avoiding Grating Lobes in radiation pattern; position of each node is ran- domly generated by Omnet++; for simplicity, only host 7 generates UDP traffic and sends it to host 0,1,3,5 that are receiver nodes. Because results de- pend on angular position and so on transmission direction of each node with respect to each other, firstly we created an utility module Mempos, to take into account the current position of the node and for generating a matrix that maintains the current posi- ion angle between all nodes, that is useful for eval- uating simulations results. As follows, we show the so called “transmission directions matrix” generated from Mempos module. It will be very helpful for results analysis:
From Fig. 8 we can note that all nodes (on re- ceiver side) receive packets. In particular, the overall received traffic is uniformly distributed among nodes. Now, we can replace the default antenna model with our PhasedArray model. We created 2 run configurations. The first one (rcl) has main lobe angle pointed to 7°, so, referring to our transmission directions matrix, we can expect that generated traf- fic will be driven towards nodes lying on main lobe region. In particular, among all receivers, only host 3 is Situated in main lobe region. Radiation pattern for Phased array with main beam angle of 7° is shown below: Figure 9: Array Factor Polar Plot for N=10, go = 7°, d = 0.52.
From Fig. 8 we can note that all nodes (on re- ceiver side) receive packets. In particular, the overall received traffic is uniformly distributed among nodes. Now, we can replace the default antenna model with our PhasedArray model. We created 2 run configurations. The first one (rcl) has main lobe angle pointed to 7°, so, referring to our transmission directions matrix, we can expect that generated traf- fic will be driven towards nodes lying on main lobe region. In particular, among all receivers, only host 3 is Situated in main lobe region. Radiation pattern for Phased array with main beam angle of 7° is shown below: Figure 9: Array Factor Polar Plot for N=10, go = 7°, d = 0.52.
Figure 12: Received Packets Omnet++ statistics (re2). Now we have maximum side lobes gain of about 2.5 dB, so side lobe power may be not fully negligi- ble in this case. Next figure shows received packets statistic among all receivers:
Figure 12: Received Packets Omnet++ statistics (re2). Now we have maximum side lobes gain of about 2.5 dB, so side lobe power may be not fully negligi- ble in this case. Next figure shows received packets statistic among all receivers:
Figure 10: Received Packets Omnet++ statistics (re1).
Figure 10: Received Packets Omnet++ statistics (re1).
In this case, data packets are mainly directed to host I which is situated within main lobe region, but due to power contribution of major side lobes, a fraction of packets will be received by host 0 also according with our powerAlgorithm. In particular, from .log analysis we observed that main side lobes present an amount power of 5 mW that is not negli- gible with respect to maximum power value in main lobe region of 10 mW. For a further validation of the model we also analyzed the results related to the energy consumption of each receiver node; for this purpose, we inserted energy simulation modules into each mobile node setting the initial energy value of each node to 150 J, and the shutdown energy value to 0 J. For Analyzing the energy consumption of nodes, we collected the Residual Capacity statistics produced by Omnet++: Figure 14: Residual Capacity rcl Config..
In this case, data packets are mainly directed to host I which is situated within main lobe region, but due to power contribution of major side lobes, a fraction of packets will be received by host 0 also according with our powerAlgorithm. In particular, from .log analysis we observed that main side lobes present an amount power of 5 mW that is not negli- gible with respect to maximum power value in main lobe region of 10 mW. For a further validation of the model we also analyzed the results related to the energy consumption of each receiver node; for this purpose, we inserted energy simulation modules into each mobile node setting the initial energy value of each node to 150 J, and the shutdown energy value to 0 J. For Analyzing the energy consumption of nodes, we collected the Residual Capacity statistics produced by Omnet++: Figure 14: Residual Capacity rcl Config..
Figure 15: Residual Capacity rc2 Config..
Figure 15: Residual Capacity rc2 Config..
Table 2: Network simulators features comparison. Although Opnet and Ns-2 provide some patches to support asymmetrical communications between nodes, they do not provide any Smart Antenna Mod- ule by default; therefore, with respect to Opnet, Om- net++ is an open source Software that can be en- hanced by developers. In this regard, our work demonstrated that, through some appropriate modi- fications, Omnet++ could be improved in terms of features, becoming more useful for end users espe- cially in beamforming and directional antenna con- texts. From Fig. 13 we can observe that the energy consumption is uniformly distributed between nodes; in Fig. 14, instead, we can note that the nodes that are not in the main lobe region preserve its en- ergy amount. In particular, all energy is conveyed toward the host that is within the main lobe region. Furthermore, in Fig. 15 the energy consumption is preserved, except from host 0 that suffers from the effect of the sidelobes. Because energy consumption results clearly reflect values obtained in packet sta- tistics, we can conclude that our designed module is functional and could be used for allowing asymmet- rical and directional communications as extension of Omnet++ features. Finally, it is shown a summary Table that compares the main features of the most used network simulators discussed in sec. 3:
Table 2: Network simulators features comparison. Although Opnet and Ns-2 provide some patches to support asymmetrical communications between nodes, they do not provide any Smart Antenna Mod- ule by default; therefore, with respect to Opnet, Om- net++ is an open source Software that can be en- hanced by developers. In this regard, our work demonstrated that, through some appropriate modi- fications, Omnet++ could be improved in terms of features, becoming more useful for end users espe- cially in beamforming and directional antenna con- texts. From Fig. 13 we can observe that the energy consumption is uniformly distributed between nodes; in Fig. 14, instead, we can note that the nodes that are not in the main lobe region preserve its en- ergy amount. In particular, all energy is conveyed toward the host that is within the main lobe region. Furthermore, in Fig. 15 the energy consumption is preserved, except from host 0 that suffers from the effect of the sidelobes. Because energy consumption results clearly reflect values obtained in packet sta- tistics, we can conclude that our designed module is functional and could be used for allowing asymmet- rical and directional communications as extension of Omnet++ features. Finally, it is shown a summary Table that compares the main features of the most used network simulators discussed in sec. 3:
Related topics:
Computer Science
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved